Health Insurance Competition
Ian McCarthy | Emory University
Do we want competition in health insurance?
Health insurance isn’t a standard market: the buyer is often an employer or the government, and the product bundles risk pooling, networks, and admin services. This begs the question, do we want to rely on standard market-based competition for health insurance?
Public Insurance:
- Simpler with less administrative overhead
- Avoids adverse selection
- Potential for wasteful spending and political influence
- Potentially lower quality
Private Insurance:
- Potentially increased efficiency and quality
- Exposure to problem of adverse selection
- Less equitable
Today’s Roadmap
- What “competition” means in health insurance
- Intermediated buyers; product = risk pooling + networks + admin
- Why risk adjustment?
- Mechanics, limits, and coding incentives
- Managed competition (Medicare Advantage)
- Bid/benchmark pricing → elasticity & markups → evidence
- What counts as success?
- Efficiency, equity, and fiscal cost tradeoffs
“Competition” in Health Insurance
Managed Competition
- The model most relevant for studying health insurance markets
- Private insurers operate under standardized, heavily regulated rules
- Examples: ACA exchanges, Medicaid managed care, Medicare Advantage
- Competition is “managed” because policy defines the terms of rivalry (risk adjustment, open enrollment, network rules)
Adverse Selection
Those with higher expected health care needs self-select into more generous coverage
- Insurers price generous policies accordingly → healthier individuals drop out (underinsurance)
- Insurers design benefits to appeal to healthy enrollees or discourage high-cost patients
Solving the problem of adverse selection
- The “easy” way: let insurers charge higher prices to sicker people → inequitable, limits access
- Managed competition instead uses risk adjustment:
- Enrollees pay uniform premiums
- Insurers receive risk-adjusted payments from CMS (or the exchange)
- Other safeguards: plan lock-in, open enrollment periods
Risk adjustment in Medicare Advantage
- Before 2000, MA attracted healthier enrollees:
- MA enrollees used ~60% as much care as FFS beneficiaries
- “Switchers” back to FFS used ~160% of average FFS spending
- CMS phased in a new risk adjustment system (2003–2006)
Challenges of risk adjustment
- Predictive power is limited (<15% of individual spending variation explained)
- But mean predictions are close → enough to reduce systematic selection
- Strong incentives for coding manipulation (“upcoding” diagnoses to raise payments)
- Design goal: reduce selection without rewarding insurers for coding intensity
Why “managed” competition?
- Insurance markets are highly concentrated
- Demand is relatively price-inelastic
- Little natural pressure for lower prices or higher quality
- → Government sets the rules of competition (benchmarks, rebates, network requirements)
Sources of market power
In Medicare Advantage (MA), 3 or fewer insurers often hold ≥95% of local market share
- High fixed entry costs (building networks, regulatory compliance)
- CMS network adequacy requirements
- Brand recognition and switching frictions
How plans are paid in MA
- Plan submits a bid, \(b\)
- CMS sets a risk-adjusted benchmark, \(B\), based on FFS spending
- If \(b < B\): plan receives \(b\) + a share of \((B-b)\) as a rebate
- If \(b > B\): plan receives \(B\) and enrollee pays \((b-B)\) as a premium
Plan pricing problem
Plans choose price/benefit generosity to maximize profit:
\[
\max_{p_{j}} \left(p_{j} + B - c_{j} \right) Q_{j}(p_{j}, p_{-j})
\]
- \(p_{j}\): plan’s “price” (premium/benefit generosity)
- \(c_{j}\): cost per enrollee
- \(Q_{j}\): demand (in risk-adjusted units)
Solution
\[
b_{j} = c_{j} + \left(\frac{d \ln Q_{j}}{d p_{j}}\right)^{-1}
\]
- Bid = marginal cost + markup
- Markup is inversely related to demand elasticity
Interpretation
- If enrollment is sensitive to benefits (elastic demand) → smaller markups
- If enrollment is insensitive (inelastic demand) → larger markups
- Evidence: a $10 bid reduction → 10% higher enrollment
- Implies markups of 10–25% above costs
Evidence from Medicare Advantage
@curto2021 study the MA market in detail:
- MA delivers care at lower resource cost than FFS Medicare
- But bids are higher than FFS costs → ~15% more expensive to taxpayers
- Distribution of “excess” payments:
- 40% → enrollees (extra benefits, lower out-of-pocket)
- 60% → retained by insurers
Key takeaway
- Market design is critical: rebates and benchmark rules shape billions in transfers
- Reforms underway: new rebate formulas, benchmark adjustments
- Bigger question: what counts as success?
How to measure success?
- Same product at lower cost than public insurer?
- Better product at the same cost?
- Costlier product, but “good enough” with higher enrollee surplus?
Conclusion: Do we want competition?
- Competition in health insurance looks very different across settings
- Small groups: insurers compete for tiny risk pools
- Large employers: insurers compete as administrators, not risk-bearers
- Managed competition: the main model, but heavily shaped by regulation
Conclusion: Do we want competition?
- Risk adjustment and managed competition can make private markets viable
- But both are fragile: selection, market power, and coding incentives remain
- Public provision avoids some problems but brings others (bureaucracy, politics)
Bottom line: Whether we want competition depends less on theory and more on how well policy design channels insurer incentives toward efficiency, equity, and fiscal sustainability.